The intelligent software paradigm is taking over surveillance due to the pervasive nature of self-learning capabilities emerging in the space. In a traditional sense, surveillance is following the movement of someone. Across the globe, governments have invested heavily in setting up a surveillance infrastructure. There are millions of cameras, but no one to monitor them 24/7. However, artificial intelligence (AI) can process every frame and present real-time analysis to save millions of lives. Numerous states are adopting sophisticated AI surveillance technologies to map, track and control people to meet various policy aims — some legitimate, some violating human rights and some of which fall into a gray area.

The reason AI is making a vast impact is because of its self-learning capabilities. Initially, the computer learned to reason about objects of interest in the scene. Now, it detects the objects in the real world, matches the results with the correct annotations (provided by humans) and tries to improve further. This way, it is constantly learning and improving. Error is minimized with every iteration and moves closer to — or even better than — human error.

Let’s discuss a few critical problems in the surveillance space. Issues like people tracking, area monitoring, parking occupancy, vehicle analytics and traffic monitoring are crucial. Amid the Covid-19 crisis, many companies have spent considerable effort in building AI-based systems to ensure social distancing in public areas. The pervasive nature of AI is leading to a level of sophistication that was unheard of a few years ago.